Research On Key Technologies Of Dental Implant Navigator Based On Near-infrared Trinocular Stereo Vision

Dental implant is a major treatment option for edentulous patients.In order to reduce operation trauma and avoid complications,dental implant navigation technology has become a widely-accepted choice.Dental implant navigation system(DINS)is able to display the operation instruments’ relative position and direction to the edentulous area in real time,thus improve operation accuracy and realize minimally invasive therapy for dental implants.Traditional DINS mainly relies on binocular stereo vision(BSV)for spatial positioning,however BSV has some shortcomings and limitations,for example:weak anti-occlusion abilities,and feature point mismatching problem.These shortcomings increase operation risk and limit the doctors’ operation scope.In order to solve or optimize these problems,near-infrared(NIR)trinocular stereo vision(TSV)has been introduced into dental implant navigation to build an DINS with high anti-occlusion ability,high precision and low cost.Components and functions of the navigator are as follows:1.T-type template:location and tracking of implant surgical instruments;2.Reference template:correction of head movement;3.U-type locating tube:oral imaging registration;4.NIR TSV:spatial positioning of optical markers;5.Navigation software:accurate guidance of implant surgery.In order to realize the above functions and improve the accuracy and robustness of the system,the following four aspects have been studied.(1)High-precision NIR camera calibration.A calibration device for NIR camera with long-term stable operation is developed,and the gray centroid algorithm based on elliptic region is proposed.The calibration accuracy of NIR camera is increased to 0.1 pixels by the high-precision calibration device and calibration points extraction algorithm.(2)Rapid extraction of optical markers and anti-mask performance.In order to ensure the real-time performance of the navigation system,the nearest neighbor fast feature point extraction algorithm is proposed.The results of TSV tracking experiments show that 60 Hz refresh frame rate can be achieved without hardware acceleration module.Aiming at the occlusion issue of optical markers in dynamic tracking,an active circular optical marker is designed and an adaptive anti-mask algorithm is proposed to guarantee the continuity and security of the system.(3)Matching and reconstruction of multi-view markers.A bipolar line constraint algorithm based on TSV ensures the uniqueness of homonymous point matching on multi-angle images.Weighted least squares method is introduced into multi-vision 3D reconstruction,and a weight estimation model is established.Reconstruction experiments show that the accuracy of this algorithm is better than that of traditional 3D reconstruction algorithm based on least squares.(4)Navigation system calibration and registration.In the process of precise guidance of oral implant surgery,the coordinate system mainly includes:the image coordinate system of oral cavity,the world coordinate system of TSV,the instrument coordinate system of implant surgical instrument,the reference coordinate system of reference template and the screen coordinate system of upper computer.In order to realize the unity of the coordinate systems,transformation relationships between the coordinate systems are established accurately through the implant instruments calibration and the image registration algorithm.In this paper,an experimental platform for oral implant navigation based on NIR TSV is developed.Localization experiments show that the accuracy of single point localization is 0.02 mm and that of tracking is 0.09 mm.According to the clinical requirements of oral implantation,phantom experiments are carried out.Results show that the mean entry deviation,exit deviation,and angle deviation are 0.55 mm,0.88 mm,and 2.23 degrees,respectively.It provides a new technical scheme for oral implant navigation.